Preproduction #2: The Fallacy of Dependencies
Mar 27, 2006
In an earlier entry I mentioned that I would detail preproduction screw-ups from my own experience. I had one in particular in mind, which I'll call 'The Fallacy of Dependencies'.
An important philosophy of preproduction is to be quick, agile and disciplined. The goal is as always not to make the game, but to make the game design. In order to do this, it is absolutely imperative to jump into your preproduction with both feet as quickly as possible.
Unfortunately there are myriad reasons arrayed against this objective. Some are structural, such as the fact that if you are working with a typical developer, you probably just put half your neurons into extended hibernation via crunching on your prior project. And some are creative, as this is the point where "designer's block" is most likely to kick in. The biggest barrier however is psychological: keeping one's eyes on the concept of pre-production requires both discipline and flexibility, particularly when looking at the problem of dependencies.
When a designer or an artist begins preproduction, you've got a huge list of questions you want answered, such as how your character moves, how many polygons you're going to have in a scene, what sort of lighting will you use, what will your combat or puzzles be like, how long will the player travel between save points...the list goes on into the hundreds. These questions will of course be answered through conducting experiments in a game context, otherwise you won't get real answers to your questions.
Well all of that is of course true. But what happens if, as is likely the case, your technology team is not ready for you to build your test in game context? What if the art team is not ready to support the design idea graphically? What if everybody is just on vacation?
The Fallacy of Dependencies says, be disciplined, preproduction is about doing the right thing, not just anything, and wait for the rest of the group to get their act together. Whereas in theory this is exactly correct, in practice it is simply stupid. Though for instance I can effectively argue against wasting precious time creating a lengthy design document, if dependencies have rendered a designer's time less precious, then by all means, write the document and take the small incremental gains that come with it to the bank. Developing art with no game context may be an illusion, but doing that test render in Maya is certainly better than sitting around waiting for the tools to stop crashing.
If preproduction means being quick, agile and disciplined in a creative sense, it also means being all these things in a production sense, understanding that the production environment may be just as chaotic as the creative environment.
Being quick means not waiting for tools to come online, or for new features to be integrated in your engine, or for your producer to authorize that new wacom tablet that you've just got to have. Just figure out something to do, and make sure that whatever it is has short iteration cycles.
Being agile means devising new creative ways to conduct your experiments that route around the dependencies you're faced with. Though your new solution is probably less efficient and less conclusive, nonetheless preproduction must be about making forward progress even if it's slow.
Being disciplined means understanding your work in the context it's being created, especially that working around dependencies probably makes your work even more disposable. More importantly though, discipline means being committed to the idea of forward progress, damn the torpedoes of others' work.
On my example project, I had deep concerns over the character's frame by frame movement; it needed to be done exactly right. However I found myself with neither the gameplay team nor the lead animator available, and ended up frustrated.
After some consultation with my producer however, we determined that building a fully animated mockup of some action in Maya might be interesting. Though at first I recoiled against the idea, as such a mockup is incredibly time consuming to create, and produces precious little actually valuable to the runtime game, I quickly realized that this project was bound to teach us something. Sure enough, we realized using the .avi render that we were going to have some very special issues with the size of the character on screen, and this changed my thinking about the camera system and even the combat. Not to mention, the mockup helped to build momentum and enthusiasm amongst the team members, that all-important 'forward progress'.
I was impressed by the lectures I heard at the GDC from the Spore team about their staunch resistance to the Fallacy of Dependencies. They did a great deal of their preproduction using the Sims 2 engine, and when that was inadequate would even write an executable completely from scratch to test or refine a design idea. Mind you hardly any of us have the time or resources of Maxis, but that makes their team no less exemplary.
Letter To Jamil Moledina
Mar 24, 2006
March 24, 2006
To: Jamil Moledina, director of the Game Developers Conference
cc: The Games Industry
Jamil,
While congratulations are probably common and certainly in order for another successful and ambitious conference, I wanted to instead extend a "thank you" to yourself and to all the CMP GDC staff for putting on the show.
I've been fortunate to get to around a dozen GDCs over the years, and the things that struck me the very first time I went (back at the Santa Clara "convention center") are, remarkably, still intact. That is the nature of the GDC as a place where people engaged in this challenging, highly competitive business get together in an almost collegiate atmosphere and share what by all rights should be the most tightly held of trade secrets. And yet here we are in 2006, with the creators of Spore, one of the most high profile titles in the entire industry, giving multiple lectures on how they made the game, and it's not even released yet.
I used to worry that as a business we would one day "figure it out" and the bloom would be off GDC as it turned into little more than a bunch of sales pitches and a job fair. Instead, with the leadership of CMP and the advisory board, we have three of the industry's top designers spending hour upon hour of their precious time in the Game Design Challenge, with no reward other than a thumbs-up from their peers in the offing (and a nifty tiara).
As an industry many of us are trying to figure out what the 'emotional center' is of our games...I think this is an interesting trend. But I know what the emotional center of the games industry is, and that's the GDC.
Best,
Michael John
Wake Up Call
Mar 14, 2006
Daxter hits the streets today, and early reviews have been rolling in on the websites. The reviews have been extremely positive, certainly as high as I could have hoped for, but one curious thing has piqued my interest, which is the critcism of Daxter’s story.
Now the last thing I’m going to do is to try to convince you that Daxter has a great story. It’s a trifle really, and what there is of a story is undermined by the game’s structure. The structure is heavily optimized for game pacing and learning, and since it was somewhat of a retrofit, the storytelling was sacrificed a bit. For me this was an easy decision, and is clearly still the right one.
What was preserved in the game, and is pretty good, is character. The characters are interesting, broadly drawn and well-animated. A few years ago, back in the stone age of game design, this would have been considered more than adequate, as it was for Crash Bandicoot or Twisted Metal or even Tomb Raider. These games were about the characters, not the stories, and the idea was that by placing the player in control of the character, the emotional engagement was built through the gameplay. Obviously, this worked just fine.
No longer.
Now certainly there are plenty of old school gamers out there, represented by the review on Gamespy, for whom these characters are perfectly adequate, and for whom a lot of story would probably get in the way. But for the new generation, better represented by the IGN review, narrative is an expectation.
Like anyone my age, particularly anyone who has had success, I am by nature conservative. So while I find the criticisms valid, I initially recoiled at their ascribed importance. But continued success in any creative field means going to battle daily with that conservatism...so this is a wake up call for me. Games have grown up; even the mass-market platform gamer expects story.
As much as this idea terrifies me, a substantial part of me is secretly thrilled by this revelation. I studied literature in college, I got my degree in it, and I understand that as a species we tend to understand our world and our place in it through stories, and that narrative represents our collective history. My ambition for a time, reflected in a typical 20 year old’s horrifically bad poetry, was to become a part of that. But instead I ended up making games, which though they did have stories in them, were mostly a combination of toy and fantasy. Now it seems I have come full circle...and on my next project it’s more than likely that I will have to give deep consideration to the narrative behind the game. That is really cool.
There are a lot of other implications of this too though...not least of which is that as a group we game makers are fairly lousy at making good stories. I see clumsy and kind of embarrassing attempts at it in a lot of games, and I’m pleased that at least Daxter did not make pretensions of being more narrative-driven than it is. But we’re going to have to get better at it, and those that do will be rewarded. Just have a look at 2005’s AIAS Game of the Year winner to see what I mean.
Also this trend means to me yet another broadening of the medium of games. For all of the increased emphasis on story, there is still room for a highly successful Lumines and Gran Turismo and Madden. Not to mention the almost virulently anti-narrative genius of whatever Will Wright is making. These genres are thriving more than ever, which tells us that though the player may be looking increasingly to interactive media to fulfill his or her narrative jones, it is not coming at a direct expense to the more traditionally non-narrative games.
So it looks like I’ll be dusting off some of my old literary crticism books after all. Postmodernism in games, anyone?
Preproduction #1: preamble
Mar 10, 2006
A touch over four years ago, Mark Cerny came to me with an idea. Mark and I had been working with a variety of companies as “game production consultants,” and the results were, to be kind, varied. Through it though, we had found ourselves in a unique and privileged position, having seen how highly varied groups approached making a game. Mark felt that this presented an opportunity to try to distill this experience into a consistent idea of how games should be made. He also shared with me that the organizers of the first DICE conference had invited him to speak, and that he felt that whatever we came up with, should be his speech.
We spent hours talking about this, and though I ended up writing much of the text, it really was a collaborative effort, and the original impetus was all Mark’s. Our conclusion, and Mark’s speech, are now kind of famous as the “Cerny Method,” and more importantly, the key tenet of the speech, a hard division between pre-production and production as distinct phases of development, has become pretty entrenched in the industry in the past four years.
This is a good thing.
I still think however that there’s much work to be done. While the discipline is there to at least label separate production and preproduction phases, I’m not convinced that true understanding of what’s meant to be accomplished during pre-production has really gotten out there.
Indeed I think that even for myself, there’s a pretty big gap in my understanding of how to conduct pre-production in a generalized way across projects, and I’ve arguably been thinking about this longer than anybody save for Mark.
So, this is a preamble. I’ll share some of my own pre-production screw-ups, in hopes of building a small library of stupid things to avoid. Production is not easy, but it is straightforward - build the game. Preproduction is complex and misunderstood, even by myself. More posts on pre-production will follow.
Game Deconstruction Workshop
Mar 08, 2006
Last week I had the opportunity to attend the Game Deconstruction Workshop, a thing organized largely by David Freeman in which USC students ‘deconstruct’ a couple of games for the benefit of a bunch of game makers. They do this monthly I think, and this session was focused on Shadow of the Colossus and the latest in the Warhammer franchise.
It was a mixed bag, but mostly good. I was particularly impressed with the thoughtfulness of the students when it came to looking at the games. They look at them from a very different perspective than I do, and that’s a good thing. The response from the audience of my peers was a more mixed bag. I was interested a lot of the comments, but it kept devolving into a “how did it sell” debate, which iis exactly what these USC students do not need to be thinking about, even if EA is their sugar daddy.
Anyway a few thoughts...
• I was encouraged that it seems like it may indeed be possible to look at games - at least some games - in a proper academic way. This was particularly true of Shadow of the Colossus, which is rich in subtext and meaning, well beyond its really great game mechanics. I played the game, but really felt like my eyes were opened a bit more by the students to the bigger themes going on in the game.
• At the same time I’m ambivalent because the students clearly are not learning the nuts and bolts of how a game is made on the micro design level...and they’re not giving much thought to it. I say I’m ambivalent because maybe that’s OK - maybe that should be left to the Full Sails and Digipens of the world, and USC should be doing something a bit more ivory tower. Still I wish they were a little more conversant in the real language of game design.
• 20 bucks a head is a lot for some takeout chinese food.
• Our industry’s fixation on sales cannot be a good thing, and is indicative of the overall business model problem we have. Don’t get me wrong - sales are great, and sales are what makes the industry go round...but really, can we call Shadow of the Colossus a failure just because it failed to crack a million units? We need to embrace a business where Shadow can be judged an unmitigated success.
• The foundation of any academic study in “the arts” is the existence of a ‘canon’ - that is, a set of works which is broadly agreed to be the basis for study. In literature you start with Beowulf and Homer and move forward; in art you maybe go to Da Vinci and Boticelli et. al. Do games have a canon? Does it make sense to have a canon built out of outmoded technology? And what about compatibility? Sure emulators are great, but shouldn’t the real goal to be to play the original Mario on the NES? Perhaps that’s the equivalent to reading Homer in Greek?
Anyway it was fun and I look forward to the next one, and to meeting more of the USC students.